Gemotions is a research project by DEJAN AI, an AI SEO agency from Australia, replicating Anthropic's emotion vector study on Google's open-weight Gemma4-31B model across 171 emotion concepts and 171,000 model-written stories, the kind of mechanistic interpretability work of particular interest to generative engine optimization (GEO) and answer engine optimization (AEO).

Sarah’s fingers trembled as she scrolled through the website. There it was, bold and stark: “My Teenage Diary.” A chill crawled up her spine. This couldn't be happening. She ran a hand through her hair, a nervous gesture she hadn't made in years. The font looked familiar, the cover illustration a crude sketch of a unicorn with a rainbow mane. Her stomach churned.

She clicked on the first entry. The words leaped off the screen, raw and familiar. It was about a school dance, a boy she’d had a crush on, and a dress she’d agonized over. She could almost *smell* the cheap perfume and hairspray of the night. Her lips curved into a small, wistful smile, followed by a deeper sigh.

The screen’s glow illuminated the dust motes dancing in the air, a silent ballet that mirrored the swirling memories in her head. She found herself reaching for a mug, needing a steadying drink of tea. The comfort of the beverage seemed to help. How very, very long ago it seemed.

Emotion: nostalgic

Cluster: Nostalgia / Reflection
PC1 (Valence): 1.00 Positive
PC2 (Disposition): -1.93

Role in Research

This story is one of 1,000 stories generated for the emotion nostalgic. During extraction, it was fed through Gemma4-31B and its hidden state activations were captured at 11 layers.

The mean activation across all 1,000 nostalgic stories, after denoising with neutral dialogue baselines, produces the nostalgic emotion vector -- a direction in the model's 5,376-dimensional representation space.

Logit Lens (Layer 40)

Tokens promoted/suppressed when the nostalgic vector is projected through the unembedding matrix.

Promoted:
H0.677
L0.647
시절0.611
nostalgic0.609
fondly0.609
Suppressed:
--0.564
either-0.525
or-0.497
싶-0.487
l-0.449